Computationally Efficient Tensor-Based MIMO Precoder

In this paper, a novel Tensor-based MIMO precoder algorithm that significantly reduces computational complexity compared to the classical SVD precoder is proposing. This approach is based on the well-known tensor decomposition (Tucker decomposition), which allows to represent and process of multidimensional data arrays efficiently. The approach significantly reduces the dimensionality of the problem leading to a substantial decrease in computational costs. Numerical experiments demonstrate that the proposed Tensor-based MIMO precoder provides comparable performance to the SVD precoder while significantly reducing computational complexity.